Antioxidant systems modulate oxidant-based signaling networks and excessive removal of oxidants can prevent beneficial acclimation responses. Evidence from mutant, transgenic, and locally adapted natural plant systems is used to interpret differences in the capacity for antioxidation and formulate hypotheses for future inquiry. We focus on the first line of chloroplast antioxidant defense, pre-emptive thermal dissipation of excess absorbed light (monitored as nonphotochemical fluorescence quenching, NPQ) as well as on tocopherol-based antioxidation. Findings from NPQ-deficient and tocopherol-deficient mutants that exhibited enhanced biomass production and/or enhanced foliar water-transport capacity are reviewed and discussed in the context of the impact of lower levels of antioxidation on plant performance in hot/dry conditions, under cool temperature, and in the presence of biotic stress. The complexity of cellular redox-signaling networks is related to the complexity of environmental and endogenous inputs as well as to the need for intensified training and collaboration in the study of plant–environment interactions across biological sub-disciplines.
The interplay between oxidant and antioxidant production is a hallmark of all living cells. While research in this area initially focused on beneficial effects of antioxidants, research in both biomedical (see, e.g., [1–3]) and botanical (see below) contexts has revised its focus due to the realization that excessive removal of oxidants can prevent beneficial acclimation responses.
While oxidation of macromolecules was initially universally viewed as damage, it is now clear that certain macromolecules are particularly sensitive to oxidation and serve as early-warning systems for rising internal oxidant production. For example, oxidation products of oxidation-prone polyunsaturated fatty acids serve as gene regulators that prompt increased antioxidant production and other responses [4,5]. Likewise, thiol-group-containing proteins are linked to redox-based gene-regulation [6–8]. Oxidation of these early-warning systems triggers enhanced antioxidant production as well as broad suites of additional adjustments in plant form and function. Antioxidant systems can thus be understood as modulators of oxidant-based signaling networks .
Plants increase antioxidant production when exposed to any of a variety of environmental stresses. The extensive body of work on this effect demonstrates antioxidant up-regulation in response to multiple stresses, which led to the assumption that engineering of crops for enhanced antioxidant production will be beneficial [10–13]. However, key questions remain. How much antioxidant production is enough? How much is too much? Does the answer to this question depend on specific internal and external contexts? What trade-offs may be associated with differences in antioxidant production and setpoints of cellular redox state?
We review evidence for mutant, engineered, and natural, locally adapted plant systems that differ in their capacity to modulate the level of oxidants formed in photosynthesis through antioxidant processes directly associated with the light-processing system of leaves. The initial impetus for this review came from observations on a pair of Arabidopsis thaliana ecotypes locally adapted to different temperature and light regimes in Italy and Sweden [14–16]. In particular, while the Swedish ecotype exhibited superior freezing tolerance [17–19], the Italian ecotype exhibited faster rosette growth and more biomass accumulation as well as lesser capacities for two chloroplast-based antioxidation processes when grown under hot temperature or low light intensity [20–25].
The following sections (i) briefly summarize the paradigm shift in how oxidants and antioxidants are viewed, (ii) provide an overview of the chloroplast-based antioxidant processes in which the Swedish and Italian ecotypes differed, and (iii) discuss different plant adaptation strategies and trade-offs associated with contrasting environments and plant growth habits as well as their implications for natural and agricultural systems.
Parallel progression in medical and plant science in understanding oxidants
Redox-signaling networks play particularly extensive roles in plants/autotrophs  that produce all of their antioxidants as needed, including tocopherols (vitamin E) and carotenoids that are the focus of this review. Furthermore, as sessile organisms, plants rely even more strongly on metabolic defenses than animals that often evade stressful environments through behavior. Stressful conditions can trigger genetic programs in plants that result in metabolic inactivation . For instance, a complete inactivation of photosynthesis in overwintering evergreens—accompanied by a precipitous drop of photochemical energy-conversion efficiency to near-zero levels—results from responses that permit evergreen leaves and needles to be retained in environments with intense solar radiation and little or no opportunity to utilize this energy for metabolism [27–31].
Earlier views on plant oxidative stress are undergoing re-examination, especially in light of the pressing need to develop climate-resilient, yet highly productive crops . For example, Mittler provided a review entitled ‘ROS (reactive oxygen species) are good’ , where he stated: ‘ROS are predominantly beneficial to cells, supporting basic cellular processes and viability, and oxidative stress is only an outcome of a deliberate activation of a physiological cell death pathway’. Foyer et al.  added: ‘This is particularly true in photosynthesis, which is driven by large redox and energy gradients, and which is the major source of ROS in plant cells. ROS production, signaling and removal associated with photosynthesis provide flexibility and control in the management of high light stress. Grasping the implications of this paradigm shift is key to addressing global issues such as food security and the production of crops in a sustainable manner for a growing world population’.
These developments in plant science parallel similar paradigm shifts in other areas. In medical science, for example, an analogous trajectory led to the conclusion that: ‘Research efforts need to be redirected. [Oxidant-based change] is protective and is a misguided target for therapy’ .
Selected examples of chloroplast-based antioxidation: de-excitation, nonphotochemical quenching, and reduction by carotenoid and tocopherol
The chloroplast's role of harnessing solar energy makes it a key site of oxidant production. The following provides a brief overview of selected antioxidant metabolites (tocopherol, or vitamin E, and carotenoids) that interact with light harvesting; excellent recent reviews are available that provide more detail ([9,36]; see also chapters in Demmig-Adams et al. ). The energy of light absorbed by chlorophyll is channeled into the electron transport chain as long as the plant is able to utilize this energy and, for example, consumes photosynthetically produced sugars in growth, reproduction, or storage. If more light is absorbed than consumed, singlet-excited chlorophyll may be converted to triplet-excited chlorophyll that is capable of passing energy on to oxygen, resulting in ROS (singlet oxygen). Excess singlet-excited chlorophyll can be pre-emptively removed by the xanthophyll zeaxanthin (Zea) in cooperation with a pH-sensing protein, PsbS [38–40]. The triplet-excited state of chlorophyll, once formed, can be de-excited by xanthophylls such as lutein . Excitation energy that makes it all the way to oxygen can be removed via de-excitation of singlet oxygen by the antioxidant tocopherol (Toco; [42,43]).
When existing antioxidant levels are insufficient to remove excess excitation energy through the above-described mechanisms, oxidant-derived regulators up-regulate genes that serve in photoprotection and other acclimatory responses. Oxidation products of polyunsaturated lipids (oxylipins) are an example of such gene regulators. Regulation of the level of singlet-excited chlorophyll affects how many electrons flow into the photosynthetic electron transport chain and thereby modulates a host of additional redox-signaling relays. These redox networks involve reduced ROS, thiol-modulated metabolites and proteins, phosphorylation cascades, and, again, oxylipins [7,9,44–46]. When excess-energy levels rise dramatically, oxidation of reaction-center proteins can shut down charge-separation as the source of high-energy electrons and incompletely reduced oxygen species [28,31,34,47]. Some authors have suggested that photoinhibition of photosynthesis is based on such an inactivation of primary photochemistry coupled with sustained engagement of photoprotective energy dissipation—as seen in overwintering evergreens (see above; [31,47–50]) or in leaves grown under constant low light and suddenly exposed to dramatically higher light intensity [34,51].
Below, we focus mainly on the first line of chloroplast antioxidant defense—pre-emptive de-excitation of excited-state chlorophyll by conversion of excitation energy to harmless heat with the help of PsbS and Zea (Figure 1). This thermal dissipation process can be monitored from the associated decreases in chlorophyll fluorescence emission, or nonphotochemical quenching (NPQ; see  and chapters in Demmig-Adams et al. ). The level of NPQ is correlated with modulation of redox-signaling pathways that regulate gene expression (see above). Moreover, both Zea and Toco have dual antioxidant functions (Figure 1)—not only in thermal de-excitation of electronically excited reactive species but also in the full reduction of incompletely reduced ROS and/or lipid-peroxidation products [52–54].
Schematic depiction of de-excitation and reduction processes in the chloroplast.
Trade-offs in the context of water use
After hearing a presentation on the proposed role of zeaxanthin in NPQ during the 1989 Rockefeller Foundation meeting on ‘The Potential of Biotechnology for Improving Grain Yield of Rice under Water Limited Conditions’ in Italy, Bill Ogren suggested that plants could be engineered with less Zea and NPQ to enhance productivity. This suggestion did not receive much interest until recently. Below, we review features of selected plant lines with impaired or naturally lower NPQ and tocopherol levels that exhibit features of interest, such as increased foliar water transport and/or increased plant growth under certain environmental conditions.
The need for climate-resilient crops is intensifying  as extreme weather becomes more common, including greater heat/drought in the summer. In his review entitled ‘Safety conscious or living dangerously’, Murchie  posed the question, ‘what is the ‘right’ level of plant photoprotection for fitness and productivity?’ This may vary dependent on the combination of specific environmental conditions and genotype. In the context of water use, annual (herbaceous, fast-growing, short-lived) plants must replenish the high levels of water lost from their leaves under hot dry conditions to sustain the high productivity needed for expedient completion of their life cycle within a single year. A greater capacity to distribute water throughout leaves is presumably coupled with other adaptations, including a larger root volume for improved underground water mining (see, e.g., . However, this strategy would not be successful when water availability in the soil decreases below a minimal level. For example, during recent decades of selection for higher yields, the overall sensitivity of corn to drought and high evaporative demand has also increased . The latter authors emphasized that the challenges of low precipitation and high evaporative demand need to be considered individually.
A review on abiotic stress tolerance entitled ‘To grow or not to grow: a stressful decision for plants’  addresses contrasting responses to drought—including accelerated growth, slowed growth, or complete growth arrest. For example, in arid and desert environments, a plethora of different strategies are employed by a variety of species, all of which are successful in their own way [58–60]. Soft-leafed annuals or desert ephemerals specialize in accelerated growth and complete their life cycle before drought sets in. In contrast, many other species arrest their growth and persist through periods of reduced water availability as evergreens or perennials that drop leaves or stems, or even persist solely as underground storage organs until sufficient precipitation returns. The approach of accelerated growth by desert annuals during a transient favorable period with high water availability may be a relevant model for irrigated annual crops with access to enough water to support continued carbon dioxide uptake under hot/dry conditions by continually replacing water lost in transpiration.
For the example of the C3 species A. thaliana, adaptations of the leaf's water-transport system have also been demonstrated. An A. thaliana accession (Col-0) originating from a site that is relatively dry for A. thaliana but not in absolute terms (558 mm annual average precipitation in the Landsberg an der Warthe region in Poland; see ) exhibited more pronounced up-regulation of transpiration rate (Figure 2A), of the density of minor leaf veins (Figure 2B), and of the proportion of each minor vein comprised of water conduits (Figure 2C) compared with an accession (Castelnuovo-12, sub-line 24) from a moister site (818 mm average annual precipitation in Castelnuovo di Porto, Italy; [14,15]) in response to growth at hot versus cool temperature under controlled conditions . In other words, an adaptation of A. thaliana to relatively low precipitation led to a greater capacity of the leaf's water-moving vascular infrastructure under hot growth temperature with high evaporative demand . This annual mesophyte's response is quite different from that of species found in much drier sites, such as woody species that restrict both water transport and growth under drought, while lessening leaf heat load via angled and/or highly reflective leaves (see discussion in ).
Phenotypic plasticity in transpiration rate and foliar water transport infrastructure in
Arabidopsis thaliana accessions adapted to different levels of precipitation.
As mentioned in the Introduction section, another set of studies reported that the A. thaliana accession adapted to the relatively mild/warm climate in Italy had a lower capacity for NPQ [21,23,25] as well as lower tocopherol levels [21,22] under certain growth conditions (including hot temperature) than an accession (Rodasen-47, sub-line 29) from Sweden (Rödåsen, Sweden; ) with its much colder climate . Future studies would be of interest that assess whether additional accessions adapted to contrasting local temperature and precipitation regimes also differ in the capacity of these antioxidants as well as, possibly, additional antioxidant systems.
Moreover, several studies on NPQ-deficient and/or tocopherol-deficient mutants of A. thaliana and other species addressed the hypothesis that these mutants exhibit leaf features that enable faster replacement of water lost in transpiration. Several A. thaliana mutants deficient in these foliar antioxidant systems (Figure 3) indeed exhibited such features when grown under hot temperature, i.e., a greater density of foliar veins and more water conduits per vein [23,61,62] that are thought to enhance the ability of leaves to replace water lost under high evaporative demand [15,63]. Consistent with these results from A. thaliana, an NPQ-deficient tobacco line with reduced PsbS and NPQ levels exhibited greater stomatal conductance, and tobacco lines with enhanced NPQ (via PsbS overexpression) exhibited lesser stomatal conductance when grown under field conditions .
Effect of antioxidant deficiency on foliar vein composition in
It may be rewarding to test whether tuning antioxidant levels to match varying water availability over the course of the growing season in regions without access to irrigation may allow plants to shift between (i) fast growth and low water-use efficiency when precipitation is available and (ii) reduced growth and high water-use efficiency during intermittent drought periods (see, e.g., ).
The A. thaliana accession from Italy that exhibited lower intrinsic NPQ capacity as well as lower tocopherol levels than the Swedish accession when grown at hot temperatures under high evaporative demand (see above) showed similarly diminished antioxidant features when grown under low light intensity [20,25]. Likewise, mutant systems with altered NPQ dynamics suggest a link between light-use efficiency and NPQ in the context of plant productivity. Tobacco engineered for accelerated disengagement of energy dissipation (accelerated NPQ relaxation) upon transition from excessively high to low light levels limiting to photosynthesis exhibited greater carbon uptake in low light as well as accumulation of significantly more biomass [66,67]. When excitation energy is no longer excessive and no other stresses are present, disengagement of energy dissipation occurs via a drop in ΔpH and removal of zeaxanthin. The tobacco leaves were engineered for either accelerated ΔpH abolishment  or accelerated removal of zeaxanthin .
It may be rewarding to address the hypothesis that this increased biomass accumulation in plant lines with lower chloroplast-based antioxidation results not only from additional available excitation energy and carbon gain in limiting light but also from possible formation of greater ROS levels and growth stimulation by oxidative signals. Future studies should quantify ROS levels in systems that return to high photochemical efficiency at different speeds subsequent to high-light exposure. Signaling processes can have profound effects on plant growth irrespective of light supply [68,69]. Future research should thus address the relative contributions to increased photosynthetic productivity from two different mechanisms: (i) reduced loss of photons via lowered thermal dissipation (i.e., reduced waste of photons) versus (ii) altered cellular redox-signaling with the outcome of maintaining, for example, stomatal opening for continued productivity under hot/arid conditions. The evidence reviewed here (e.g., that both NPQ-deficient mutants and tocopherol-deficient mutants exhibit increased support for foliar water transport when grown at hot temperature) points to the importance of redox modulation. While NPQ-impaired mutants concomitantly lower thermal dissipation of absorbed photons and also impact cellular redox state, the tocopherol-deficient mutant featured here exhibited unaltered thermal-dissipation capacity (see Stewart et al.  for NPQ capacity in hot-grown vte1), and presumably acts exclusively through its impact on cellular redox state. The similar effect of these two mutant systems on foliar water transport (Figure 3) suggests redox modulation as the common mechanism for this effect.
Growth rates did differ between the A. thaliana accession from Sweden with the higher NPQ capacity and higher tocopherol levels and the Italian accession with the lower NPQ capacity and lower tocopherol levels in low-light- or hot-grown leaves. The Swedish accession produced much smaller rosettes and less biomass than the Italian accession under either low growth light intensity or hot growth temperature [20–22,70]. While this finding supports a hypothesis that lesser NPQ/tocopherol levels may be associated with growth stimulation by oxidative signals under certain environmental conditions, it is important to keep in mind that these two natural accessions differ in multiple regulatory networks (e.g., ), of which redox-regulatory systems are just one component.
Trade-offs in the context of temperature
The evidence discussed above suggests that rapid replacement of lost water under hot conditions in annual C3 species via greater foliar vein density and more water conduits per vein may be associated with a shift of the cellular redox state to more oxidizing conditions. The question arises whether the converse would be true for cool conditions, i.e., that a shift of the cellular redox state to more reducing conditions may be beneficial. Identification of possible trade-offs in plant performance in hot–dry versus cool-moist environments is important since the current increase in extreme weather events includes not only greater heat and/or drought, but also late-spring/early-fall cold spells (Intergovernmental Panel on Climate Change; http://www.ipcc.ch).
While greater intrinsic NPQ capacity and tocopherol levels in the Swedish versus the Italian accessions of A. thaliana were reported specifically for plants grown under low light intensity and/or hot temperature (see above), no such data are available for plants grown under high light and/or cold temperature. Future characterization should thus be extended to additional environmental growth conditions. In the meantime, there is indirect evidence supporting the hypothesis that the Swedish accession may, in fact, produce greater rather than lesser concentrations of oxidant-derived oxylipin gene regulators in either cold- or high-light-grown leaves than the Italian accession. This indirect evidence stems from three lines of inquiry detailed below on cell wall ingrowths in foliar sugar-loading cells as affected by genotype and experimental oxylipin treatment.
Cool temperature diminishes plant productivity by lowering the activity of proteins, including enzymes involved in carbon fixation  and transporters involved in sugar export from leaves. Expedient sugar export from leaves is important to maintaining high levels of photosynthesis because sugar build-up in leaves can activate feedback that leads to repression of photosynthetic proteins (see ). The Swedish accession exhibited greater up-regulation of intrinsic photosynthetic capacity (Figure 4A) as well as of the number (Figure 4B) and cell wall invaginations of phloem cells involved in loading sugar into export conduits (Figure 4C) in cool- versus warm-grown leaves than the Italian accession . Both greater numbers and wall ingrowths [22,73] of these phloem cells enhance the membrane area available for placement of transport proteins that support sucrose loading (see discussion in [16,24]).
Phenotypic plasticity in photosynthetic capacity and foliar sugar-transport infrastructure in
Arabidopsis thaliana accessions adapted to different temperatures.
It was, furthermore, shown that treatment with jasmonate (methyl jasmonate, MeJA) enhanced wall invagination of phloem cells in A. thaliana (Figure 5) . A similar enhancement was seen in an NPQ-impaired A. thaliana mutant compared with its wild-type (WT) (Figure 5) . These results support the hypothesis that less photoprotection, and resulting enhanced oxylipin production [53,75,76], could be beneficial under cool growth temperature in annual species. Future studies into the antioxidant capacity of cool-grown leaves should address a broader variety of antioxidant processes and their possible differential adjustment in response to environmental conditions.
Effect of the phytohormone methyl jasmonate or antioxidant deficiency on foliar sugar-loading cell ultrastructure and effect of antioxidant deficiency on lipid-peroxidation-based messengers.
Vascular tissue is an attractive target of jasmonic acid (JA) signaling since JA is not only synthesized in the plastids of phloem cells but is also transported over long-distances through the phloem [77–79]. Future studies should address whether JA may also be involved in the adjustment of water conduit number and vein density shown in Figure 3. Overall, oxylipin gene regulators should be a rewarding target of future investigation into the impact of chloroplast antioxidant capacity on plant growth, development, and stress response, since enzymatic lipid peroxidation by lipoxygenase (LOX) is activated via oxidation of LOX's catalytic iron center by ROS and inactivated via reduction of this iron by tocopherols . Other hormones, like auxins that are known to orchestrate vascular differentiation , may also be involved since biosynthesis, degradation, and signaling of multiple plant hormones is under redox control [82,83].
In addition to evaluating both sugar- and water-transport components of the foliar vasculature, future studies should consider specific features of individual water conduits that protect against tension-induced conduit collapse under high evaporative demand (thicker cell walls; see, e.g., [84,85]) or against freeze–thaw-induced embolism (narrower conduits; see, e.g., [86–88]). Both tocopherol-deficient and NPQ-deficient A. thaliana mutants exhibited an intriguing combination [23,61,62] of greater vein density and greater water-conduit numbers (as features with putative benefits under heat/drought) with narrower water conduits (as a feature with putative benefits under freeze–thaw cycles).
Beyond abiotic conditions: biotic defense, NPQ, and tocopherol
In the context of effects of redox state on foliar vascular organization illustrated by Figures 2 and 3, it should be noted that many pathogenic viruses, bacteria, and fungi spread through the plant vasculature [89–91]. Modulation of foliar vascular organization, among other defense-related responses, should thus receive further attention as a target of signaling cross-talk and cross-tolerance of abiotic and biotic stress. A PsbS-deficient mutant of rice exhibited greater resistance against fungal and bacterial pathogens . Similarly, the PsbS-deficient npq4 mutant of A. thaliana exhibited superior defense against herbivory by caterpillars  and spider mites  compared with WT. It should be noted, however, that seed production was reduced in the PsbS-deficient npq4 mutant of A. thaliana under field conditions . To identify possible trade-offs, a comprehensive characterization of biotic and abiotic stress responses is needed under multiple environmental conditions and in multiple genotypes.
In addition to their role in biotic defense, chloroplast-based redox-signaling networks can themselves be the target of invaders . For example, the necrotrophic pathogen Sclerotinia sclerotiorum causes a decrease in intra-thylakoid pH associated with stimulation of Zea and NPQ formation as well as a drop in violaxanthin that is not only a precursor of Zea but also of the plant hormone abscisic acid (ABA) with its roles in stomatal closure and defense responses . Increased stomatal opening facilitates pathogen entry into the leaf and compromised defense further facilitates infection . This pathogen thus manipulates Zea and thermal dissipation levels to enhance its ability to infect plants. These results further highlight the need for comprehensive studies into the impact of antioxidant-capacity modulation under multiple environmental challenges.
Redox-signaling networks mirror complexity over space and time in environmental and endogenous contexts
The remarkable complexity of redox-signaling networks mirrors that of whole-plant response to a wide range of specific environments with myriad combinations of abiotic and biotic influences [60,98]. Environmental cues are perceived and integrated by transduction systems involving the interaction of ROS, antioxidants, phytohormones, and sugars with transcription factors that modulate stress tolerance, programmed cell death, growth, and development [45,99]. ROS can either promote or suppress growth by having different effects on cell cycle and programmed cell death and depending on the specific type of ROS and tissue involved, plant developmental state and source–sink balance, and the kind of environmental stress encountered [33,100].
Rather than representing a single optimal setpoint that provides homeostasis of metabolism, oxidant/antioxidant-balance setpoints may differ dependent on the external environment and genetic differences in adaptation to local environmental conditions. Shifts in these setpoints may have occurred over the course of plant evolution and also take place during plant development.
Esteban et al.  proposed an evolutionary shift from greater emphasis on pre-emptive de-excitation of excess singlet-excited chlorophyll via Zea-associated NPQ to a greater emphasis on detoxification of ROS once formed by tocopherol, which could be seen as shifting from being more ‘safety conscious’ to ‘living dangerously’ (see above; ) as terrestrial environments and climate became more variable. A follow-up hypothesis could be formulated stating that the relative emphasis also differs between fast-growing herbaceous and slow-growing woody species with their very different intrinsic NPQ capacities (i.e., determined under experimental conditions that allow only enough electron transport to maintain trans-thylakoid pH) .
Moreover, shifts in photoprotection also occur during the progression through plant life stages. For example, Juvany et al.  reported that females of a Mediterranean tree species lowered both NPQ and tocopherol levels and increased LOX activity, all of which presumably increased oxylipin production, when entering their reproductive phase. Such a down-regulation of antioxidant capacity may serve to enhance the production of oxidant-based orchestrators of reproduction.
Conclusions and outlook
The existing evidence indicates that both environmental and endogenous conditions shift the setpoint for the balance between oxidation and antioxidation. Further work is needed that places photoprotection via chloroplast-based antioxidant processes in the context of redox-signaling networks controlling plant growth, development, and stress response . Such investigation should be conducted in whole plants  under a variety of environmental conditions and with multiple genotypes. Specific environmental conditions should include both abiotic and biotic factors and their variation over the course of days and seasons in the context of different plant genetic and developmental features (see ). These comprehensive approaches will require intensification of discipline-transcending training and collaboration across molecular science, genetics/genomics, functional biology, ecology, and evolutionary biology.
We thankfully acknowledge support from the National Science Foundation [award DEB-1022236] and the University of Colorado at Boulder.
The Authors declare that there are no competing interests associated with the manuscript.